Measuring change in alcohol
brief intervention trials: How are
consensus core outcome sets
developed?
Dr Gillian W. Shorter
and the ORBIT Team (PI: Dorothy Newbury-Birch)
Funding: Alcohol Research UK
The challenge of complexity
• “Trials are only as credible as their outcomes”
(Tugwell 1993)
• Beyond trials
– Funding
– Reviews
– Understanding
• Service user views
Why is it needed?
• Raise hand if you have done a meta analysis
on any topic related to alcohol
• How much of the data could you use from the
papers in your meta analysis?
• We use different measures, and this affects
the potential to synthesise the evidence
(inconsistency problem)
Systematic review evidencing
outcome reporting bias
• The problem of significance
– Trial more likely to be published
– Outcomes more likely to be fully reported
• 40–62% of publications had 1+ primary
outcome changed, newly introduced or
omitted compared to protocol
[Dwan et al, PLoS ONE 2008]
4
Core outcome set
• Consider both benefits and harms
• The minimum (other outcomes can be
collected)
• Focus is on trials of effectiveness
• “What” to measure, then “How”
How ORBIT will approach this
complexity is crucial
The masterplan
Systematic
Review
• What
• How
Delphi
• Two round
• Consensus
meeting
Dissemination
• Publication
• Guidance
Why INEBRIA is important
• We need your input
– Delphi
– Outcome challenges
– Concerns
– Please get in touch with me (and can be
confidential)
[email protected] [email protected]
An example from OMERACT: What
do we gain from COS generation
Before….
Tender joints
Swollen joints
10%
Patient global
Physical
measured
assessment
disability
all of these
Now (ish) Physician
Pain
global
assessment
70%
Acute
phase
Radiological
measured
reactants
change
all of these
What it might look like
Zarin et al 2011; DOI: 10.1056/NEJMsa1012065
Advantages of core outcome sets
• Increases consistency across trials
• Strengthen evidence base
• Much more likely to measure
appropriate outcomes (stakeholders)
• Major reduction in selective reporting
11
In conclusion
• Benefits: What is meaningful change, and to
whom does it have meaning?
• Please get in contact
– Take part in the Delphi (both waves), give me your
views (good, bad, indifferent, where outcomes are
useful for you)
• [email protected] [email protected]
With thanks to Alcohol Research UK, COMET initiative management group, COMET
Initiative, Dorothy Newbury-Birch, Nick Heather, Emma Giles, Amy O’Donnell,
Carolina Barbosa, Aisha Holloway, Mike Clarke, Jeremy Bray, and the INEBRIA
Outcome Reporting in Brief Intervention Trials SIG